Production networks in Europe: A natural experiment of the European Union enlargement to the east

• DOI: http://doi.org/10.1111/roie.12484
• Published date: 01 November 2020
• Author: Martina Vidovic,Inmaculada Martínez‐Zarzoso,Anca M. Voicu
• Date: 01 November 2020

Rev Int Econ. 2020;28:1143–1163.

|

1143

wileyonlinelibrary.com/journal/roie

1

|

INTRODUCTION

Proximity plays an important role for trade relations, in particular, when the products traded are

intermediates that are used in several stages of the production process. In addition, the size

of the market and relative factor endowments or economic behavior are equally influential

DOI: 10.1111/roie.12484

ORIGINAL ARTICLE

Production networks in Europe: A natural

experiment of the European Union enlargement to

the east

InmaculadaMartínez-Zarzoso1,2

|

Anca M.Voicu3

|

MartinaVidovic3

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution

and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

© 2020 The Authors. Review of International Economics published by John Wiley & Sons Ltd

1Department of Economics, Georg-August

University of Göttingen, Goettingen,

Germany

2Department of Economics, University

Jaume I, Castelló, Spain

3Department of Economics, Rollins

College, Winter Park, FL, USA

Correspondence

Inmaculada Martínez-Zarzoso, Department

of Economics, Georg-August Universitaet

Goettingen, Platz der Goettinger Sieben 3,

Goettingen 37073, Germany

Email: martinei@eco.uji.es

Funding information

Ministry of Economy

Abstract

This paper focuses on the 2004 enlargement of the European

Union to the East and treats it as a natural experiment to in-

vestigate two issues: first, whether there has been a trade

creation effect in final and intermediate goods and sec-

ond, to what extent this effect has been more pronounced

for final or for intermediate goods. Using difference-in-

differences analysis, we find that the effect of the 2004 EU

enlargement has been positive for both intermediate and

final goods trade, and it is in general greater for final goods.

Using a generalized gravity model of trade that controls for

the multilateral resistance and bilateral time-invariant fac-

tors, we estimate an increase in bilateral trade of 28% for

final goods and 24% for intermediates. However, the effects

are heterogeneous by sub-sector and indicate that the main

trade gains were for non-durable consumer goods and food

and beverages primary and processed products.

JEL CLASSIFICATION

F14

1144

|

MARTÍNEZ-ZARZOSO ET Al.

(Egger, Pfaffermayr, & Schmidt,2007). International trade in parts and components has intensified

in the last decades as a result of increasing globalization and reducing trade costs. This develop-

ment has been of special relevance in East Asia, starting in the 1990s, but also to some extent in

Europe. Several authors investigated the importance of trade and production networks as a means

of fostering economic growth and development in Asia (Kimura, Takahashi, & Hayakawa,2007;

Ng & Yeats,2001) and in Europe (Egger & Egger,2005; Feenstra,1998). Other authors focused

on the welfare effects of regionalism (Egger & Pfaffermayr,2013; Frankel, Stein, & Wei,1996;

Krugman,1991a, 1991b) and t hose of the European Agreements including an assessment of the

trade effects, but without distinguishing between trade in final and intermediate goods (Caporale,

Rault, Sova, & Sova,2009; Egger & Larch,2011; Spies & Marques,2009). Spies and Marques’

(2009) paper uses an augmented gravity model that rely on the inclusion of country pair fixed ef-

fects in order to avoid biases found in previous studies. The specification used by the authors leads

to evidence that the free trade agreements (FTAs) with the Central Eastern European Countries

(CEECs) have led to significant increases in intra-CEECs trade. Their results show that once the

omitted variable bias was corrected, the FTAs with the CEECs resulted in 7% to 20% more new

trade compared to the scenario that included only time-invariant country pair effects.

Since the second half of the 2000s a few attempts have been made to investigate whether, as in

Asia in the 1990s, a similar pattern of emergence of production fragmentation has occurred on the

European continent, and, in particular, following the accession of the CEECs into the European Union

(EU) (Curran & Zignago,2012; Kaminski & Ng, 2005; Kaplan, Kohl, & Martínez-Zarzoso, 2017;

Martínez-Zarzoso, Voicu, & Vidovic,2015; Zeddies,2011; among others).

It is no surprise that Western and Eastern European countries are considered natural trading part-

ners due to their proximity and historical ties. Indeed, the Europe Agreements in the early 1990s

already established bilateral free trade between the EU and each individual CEECs in most industrial

products (Egger & Egger, 2005). However, a number of artificial trade barriers, different from tariffs

and non-tariff barriers still remained. Namely, behind-the-border trade barriers such as administrative

burdens or differences in products standards that deter international trade to a non-negligible extent

(Hornok,2010; Wilson, Mann, & Otsuki, 2003, 2005). Since tariffs and non-tariff bar riers were al-

ready eliminated in the 1990s, the accession1 of eight2 CEECs into the EU in 2004, of Romania and

Bulgaria in 2007 and of Croatia in 2013 provides a quasi-natural experimental setting that can be used

to investigate the importance of behind-the-border barriers across integrated markets (Hornok,2010).

In particular, this setting can be used to infer whether these barriers affect intermediates and final

products differently. According to the theory of fragmentation (Jones, Kierzkowski, & Lurong,2005)

income and trade cost variables prove to be important in affecting the magnitude of trade in interme-

diates and final goods. Specifically, these variables are expected to have a stronger impact on trade

in parts and components than on trade in final goods. In this study we use a difference-in-differences

(DID) strategy and estimate a generalized gravity model of trade to investigate the effects that the ac-

cession of the CEECs into the EU had on bilateral trade of final and intermediate products.3 We also

investigate whether our results support the abovementioned theory.

To our knowledge this is one of the first papers that show evidence of the effect of the 2004 EU

enlargement for bilateral trade in intermediate and final products separately. We depart from Kaplan

et al. (2017) in that we use different trade data. The data in our paper come from trade statistics and

not from input-output tables (IO). Value added trade figures based on IO rely on technical coefficients

that are only available for some years, hence some figures are interpolated. In addition, our data cover

a more recent period and provide more time variability. We restrict the data to trade in manufactured

goods and, hence, trade in services is not considered, given the heterogeneity of the services sector

and the lack of data for some of the countries considered.